[MLton] Windows port of MLton using the Microsoft tools (e.g. without MinGW)

[Nicolas Bertolotti]

In response to < how do you mean < does not require it anymore > >, I guess
I should have written "would not" instead. Sorry, I'm french and my English
is not so good.

 

The reason why a "real" 64 bits version is interesting for me is that my
software uses a lot of RAM and the more it can address, the better it is.

 

There are 3 reasons why I'm considering porting the MLton compiler. 

 

The first one is that I'm afraid it would be hard for me to contribute to
"gcc" whereas it is much easier for me to contribute to MLton. 

 

The second one is that the Windows version of GNU tools is usually available
much later than the Linux version (probably because the GNU community does
not like Windows) and this is exactly what's happening here. The 64 bits
version of Windows has been released a long time ago and the GNU tools still
don't support it.

 

The third one is that another compiler for an ML language (Objective Caml)
which is available for both MinGW and the Microsoft tools already supports
Win64. As a consequence, I thought it could be the way forward.

 

  _____  

From: Wesley Terpstra [mailto:wesley at terpstra.ca] 
Sent: Thursday, July 26, 2007 2:54 PM
To: Nicolas Bertolotti
Cc: mlton at mlton.org
Subject: Re: [MLton] Windows port of MLton using the Microsoft tools (e.g.
without MinGW)

 

On Jul 26, 2007, at 1:17 PM, Nicolas Bertolotti wrote:

I'm interested in getting a Win64 version of MLton. Now the 64 bits port for
Linux (and also other targets as I could read some messages about it on the
mailing list) is getting stable, I guess it should not be too complex to
update the MinGW port of MLton for Win64.

In contrast to skaller, I think this is the way forward. MinGW is a very
thin wrapper that provides most of what a UNIX-ish app like MLton needs. As
you can still link to and FFI to MS compiled code, there's little downside.



Unfortunately, MinGW is not yet available for Win64. The gcc port is
definitely not stable (the binutils seem to run fine) and the MinGW headers
and libraries are still experimental.

Are you in a hurry? I'm sure that gcc for windows will work in the near
future. Fixing any problems in the MinGW headers once gcc works seems like
much less work than trying to port MLton and its dependencies to MS's
compiler.

The fact we need MinGW to use MLton on Windows seems to be an obstacle in
terms of portability and the fact it does not require it anymore would make
life (at least mine ;-)) easier.

How do you mean "does not require it anymore"? I'd not heard of anyone
porting MLton to a non-gcc compiler?

 

In the meantime, 32-bit applications still work great (and use less memory).
So unless need >2GB of memory in a MLton app, you're not missing much.

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mlton.org/pipermail/mlton/attachments/20070726/e8f64f2d/attachment.htm